Effect Of Klex Decentration On Postoperative Corneal Higher-Order Aberrations After Uneventful Smile: A Strictly Matched Cohort Analysis

SMILE, as the first KLEX procedure, is well established, with more than 13 million procedures performed worldwide. Postoperative visual outcome, however, does not always match the intended results. Accurate centration during docking remains a critical step. This study aimed to quantify the effect of KLEX centration, measured as the vector offset from a Purkinje reflex based visual axis proxy, on postoperative corneal higher-order aberrations after uneventful SMILE. The goal is to better characterize the sensitivity and forgiveness of decentration and to assess impact of contributing ocular factors.

Retrospective cohort study including consecutive patients who underwent uneventful SMILE between 2014 and 2025 at the Department of Ophthalmology, Ludwig Maximilian University of Munich, and the SmileEyes eye clinics in Munich, Germany. Eventful procedures and insufficient follow-ups were excluded. Strict matching addressed surgeon, VisuMax Platform laser delivery settings, optical zone, and postoperative medication. stratification was based on preoperative subjective sphere and astigmatism.

The desired centration point as visual-axis proxy was derived from preoperative IOLMaster images by extracting the x–y coordinates of the first Purkinje reflex. For intraoperative centration assessment, images captured during VisuMax docking were analyzed through ImageJ software. The displacement of the docking/treatment center from the pupil center was measured and converted to a vector characterized by magnitude and angular direction. Pentacam HR measurements acquired ≥3 months after surgery were evaluated at a 6.0-mm analysis diameter. Assessed HOAs included total HOA RMS (3rd–6th), Z4^0 spherical aberration, Z3±1 coma (0°/90°), and Z3±3 trefoil.

A total of 288 eyes were analyzed across six matched subgroups. The mean KLeX decentration distance was 0.28 ± 0.13 mm. A significant correlation was observed between decentration distance and induced total coma (r = 0.64, p < 0.001). The most affected HOA was vertical coma, followed by spherical aberration and trefoil. Vector decomposition showed a stronger association for the vertical than the horizontal decentration component. A dose–response pattern (threshold) effect was observed, with a steeper increase in HOA in higher astigmatism strata, indicating reduced centration tolerance.

A decentration forgiveness diagram was generated to visualize tolerance ranges across astigmatism strata.

Centration is a fundamental determinant of outcomes in corneal refractive surgery. Quantifying its impact can support risk-adapted clinical workflow prioritization. The proposed image analysis sequence reliably quantified decentration vector and could be integrated into the operative summary to document centration Quality and improve possible enhancement planning.